The mathematical text in its published form, as we are most used to reading it, is a carefully structured and polished means of communicating results to the scientific community. It is, as Reuben Hersh put it, the 'front' of mathematics. In this paper, I propose to look at the 'back' of mathematics, at what happens in the privacy of drafts, which can certainly be seen as the mathematician's laboratory. Considering that these preliminary texts are a part of the mathematical practiceand indeed a crucial one -I will show that they allow us to understand the shaping of mathematics in deep and significant ways. Using a selection of examples, I will focus on questions related to the materiality of mathematical texts, how textual elements and mathematical practices work with each other, the processes of writing in mathematics, and the choices made in writing a text deemed suitable for communication to the scientific community. 1 [Medawar(1991)]